Mutational K-ras activation arises very frequently in epithelial cancers, with highest frequencies prevalent in lung, pancreatic and colorectal cancers. K-ras mutant cancers are highly aggressive with extremely poor prognoses and high mortality rates. Unfortunately, these cancers remain highly refractory to both conventional and targeted therapeutics. Additionally, pharmacological inhibition of K-Ras protein has proven to be difficult to achieve clinically. My initial aim was to model "addiction" to mutant K-ras in the hopes of identifying molecular determinants that define the "Kras-addicted" state with the ultimate aim of identifying candidate therapeutic targets. Using K-ras-specific shRNAs to deplete K-Ras protein in lung and pancreatic cancer cell lines harboring mutant K-ras, I identified two classes - those that are "addicted" to or dependent on K- Ras activity for viability and those that are K-Ras independent. By comparing whole genome expression profiles for these two classes, I derived a gene expression signature associated with K-Ras dependency. Several genes represented in this signature were required to maintain the viability of K-Ras dependent cancer cells, implicating them as candidate therapeutic targets in Kras mutant cancers. This proposal hopes to extend these initial findings by identifying lineage or tissue-specific anti-apoptotic genes in Kras mutant cancers, with the hopes of revealing novel context-specific therapeutic targets. Tissue-specific K-Ras dependency signatures will be generated for Kras mutant lung, pancreatic and colorectal cancers. Genes representing pharmacologically tractable targets will then be validated for their potential as candidate therapeutic targets. Cell biological and biochemical analyses will be performed to determine if there are tissue-specific biological or signaling features that contribute to the general phenomenon of K-Ras dependency. Finally, available pharmacological inhibitors against these targets will be tested for efficacy in pre-clinical mouse models of K-ras-driven tumorigenesis. PUBLIC HEALTH RELEVANCE: Kras is the most frequently mutated oncogene in solid tumors with highest frequencies seen in lung, pancreatic and colorectal cancers. I propose to generate gene expression signatures for lung, pancreatic and colorectal cancers in the hopes of identifying novel therapies for each of these cancers, which are very difficult to treat with current chemotherapeutics.